F-GRPO factorizes group-relative policy optimization into generation and ranking phases within one autoregressive sequence, using order-invariant coverage and position-aware utility rewards to improve top-ranked performance on recommendation and multi-hop QA tasks.
Title resolution pending
36 Pith papers cite this work, alongside 739 external citations. Polarity classification is still indexing.
36
Pith papers citing it
739
external citations · Crossref
representative citing papers
dGRPO merges outcome-based policy optimization with dense teacher guidance from on-policy distillation, yielding more stable long-context reasoning on the new LongBlocks synthetic dataset.
SIOP enables turn-level credit assignment in LLM agents via semantic clustering of final answers as latent outcomes, improving performance on reasoning benchmarks without verifiers.
SENECA uses a novel self-consistent missing mass calculation to improve discrete entropy estimates in small-sample regimes and outperforms alternatives in numerical tests.
InvEvolve evolves white-box inventory policies from LLMs with statistical safety guarantees and outperforms classical and deep learning methods on synthetic and real retail data.
ReaLM-Retrieve uses step-level uncertainty to trigger retrievals during reasoning, achieving 10.1% better F1 scores and 47% fewer calls on multi-hop QA benchmarks.
SOB benchmark shows LLMs achieve near-perfect schema compliance but value accuracy of only 83% on text, 67% on images, and 24% on audio.
Fine-tuning shows higher proficiency than in-context learning on in-distribution generalization in formal languages, with equal out-of-distribution performance and diverging inductive biases at high proficiency.
Language models frequently violate temporal scope stability in multi-turn dialogues by drifting toward present-day assumptions even when they possess the correct facts.
Abstract-CoT lets models reason with short discrete latent token sequences from a reserved vocabulary, using warm-up training and RL to match verbal CoT performance with up to 11.6x fewer tokens.
Semantic stratification organizes documents into entity-based clusters to systematically generate queries for missing strata, yielding formal coverage guarantees and interpretable failure mode visibility in retrieval evaluation.
PTR framework profiles a workflow upfront then executes it deterministically with bounded verification and repair, limiting LM calls to 2-3 while outperforming ReAct in 16 of 24 tested configurations.
M3-Embedding is a single model for multi-lingual, multi-functional, and multi-granular text embeddings trained via self-knowledge distillation that achieves new state-of-the-art results on multilingual, cross-lingual, and long-document retrieval benchmarks.
A²TGPO improves RL policy optimization for multi-turn agentic LLMs by normalizing information gain within same-depth turn groups, rescaling cumulative advantages by sqrt of term count, and modulating clipping ranges per turn's normalized IG.
Attention-based models can retrieve evidence intrinsically by using decoder attention to score and reuse their own pre-encoded chunks, outperforming separate retrieval pipelines on QA benchmarks.
SANTA sparsifies post-softmax value aggregation via stratified sampling of S << n_k indices to produce an unbiased estimator, delivering 1.5x decode attention speedup on RTX 6000 Ada at 32k contexts while matching baseline accuracy.
Verbal-R3 uses a verbal reranker to generate analytic narratives that guide retrieval and reasoning in LLMs, achieving SOTA results on complex QA benchmarks.
A survey that categorizes RIR benchmarks by domain and modality, proposes a taxonomy for integrating reasoning into retrieval pipelines, and outlines key challenges.
CL-bench Life shows frontier language models achieve only 13.8% average success on real-life context tasks, with the best model at 19.3%.
R³AG routes queries to retrievers by decomposing capabilities into retrieval quality and generation utility, trained via contrastive learning on document assessments and downstream answer correctness to outperform static methods.
GRIL uses stage-specific RL rewards to train LLMs to detect missing premises, pause proactively, and resume grounded reasoning after clarification, yielding up to 45% better premise detection and 30% higher task success on insufficient math datasets.
KnowledgeBerg benchmark shows open-source LLMs achieve only 5.26-36.88 F1 on universe enumeration and 16-44% accuracy on knowledge-grounded compositional reasoning, with persistent failures in completeness, awareness, and application.
Joint RL training of reasoning agent and document ranker via GRPO with semantic grouping and composite rewards yields consistent gains over fixed-retrieval baselines on seven QA benchmarks.
NWCAD uses a two-stream setup with a two-stage gate to prevent accuracy drops on baseline-correct items under non-informative contexts while retaining gains from helpful contexts.
citing papers explorer
-
F-GRPO: Factorized Group-Relative Policy Optimization for Unified Candidate Generation and Ranking
F-GRPO factorizes group-relative policy optimization into generation and ranking phases within one autoregressive sequence, using order-invariant coverage and position-aware utility rewards to improve top-ranked performance on recommendation and multi-hop QA tasks.
-
Self-Induced Outcome Potential: Turn-Level Credit Assignment for Agents without Verifiers
SIOP enables turn-level credit assignment in LLM agents via semantic clustering of final answers as latent outcomes, improving performance on reasoning benchmarks without verifiers.
-
InvEvolve: Evolving White-Box Inventory Policies via Large Language Models with Performance Guarantees
InvEvolve evolves white-box inventory policies from LLMs with statistical safety guarantees and outperforms classical and deep learning methods on synthetic and real retail data.
-
Retrieval from Within: An Intrinsic Capability of Attention-Based Models
Attention-based models can retrieve evidence intrinsically by using decoder attention to score and reuse their own pre-encoded chunks, outperforming separate retrieval pipelines on QA benchmarks.
-
Stochastic Sparse Attention for Memory-Bound Inference
SANTA sparsifies post-softmax value aggregation via stratified sampling of S << n_k indices to produce an unbiased estimator, delivering 1.5x decode attention speedup on RTX 6000 Ada at 32k contexts while matching baseline accuracy.
-
Benchmarked Yet Not Measured -- Generative AI Should be Evaluated Against Real-World Utility
Generative AI evaluation must shift from static benchmark scores to measuring sustained improvements in human capabilities within specific deployment contexts.